Breaking new Grounds (An Engineering Exploration in Building a Coffee Machine, part II)
With a lofty vision to produce a great cup of coffee I quickly ventured into a place between trepidation and excitement. I remind myself this is where growth happens, suspend disbelief, and begin questing for parts. I started the quest by searching for what I called “anchor-parts”. Anchor-parts, by my definition, would be the core parts that are likely to remain unchanged throughout design process. I model these first, and begin designing the more advanced components second.
While I was designing for function and form, it became increasingly common to build something, and rework it into something better. Function and aesthetics were tidal-locked with my ability to use the design tools. I would need to up-level my abilities with the tools (drafting, modeling, placement, integration), and my mechanical design knowledge (torque forces, bolts, materials, mounts) to unlock iteration on the aesthetics (look, feel, shape, holistic form).
As I poured over the sketches, and engineering drawings, the research projects took on a life of their own and started to become more tangible. Vision was turning from concept-art into sub-assemblies, mock-prototypes, and brew tests. With a hybrid set of brew methods, manual tests, half functioning prototypes, custom filters, tweaks, design, software updates, electronics, I would eventually have a full working prototype.
For each of the goals defined in part I, I have included a brief summary and notes on how I achieved each goal.
It needs to make coffee consistently, and repeatably.
- In order to achieve this, I built several mini prototypes, to test the basic mechanics, electronics, taste of coffee and continued to refine and enhance the design along the way. The final machine comprises a combination of custom designed and hand-machined parts, off-the-shelf parts, printed parts, as well as several precisely machined CNC parts to achieve the desired goal.
- Throughout each iteration, I upgraded each part to higher and higher quality components, from sensors, to precision controls, better power supplies, higher quality glass, stainless steel, aluminum, waterproof enclosures and made countless upgrades to the software control.
- Given the same grounds, the same temperature, the same amount of water, the taste is consistent and repeatable. It handles the temperatures of the brew cycle, is relatively quiet to operate and has now made many single cups of coffee.
- Notes: I would have to learn about the mechanical and electrical components required to move custom machined parts precisely. Other areas I needed to learn included, sensors, valves, motors, control boards, torque application requirements, power supplies, and software development workflows for the hardware modules (from new IDE’s to available libraries to the bootstrapping/real-time run control). There would be some pieces that could only be done via CNC. My general feeling was: there’s a lot here, but I’ll figure it out, piece by piece.
It needs to have an aesthetic that I would appreciate on my counter (I’m a big fan of Apple® design and engineering).
- Refining the design several times, while learning new techniques to draft and machine parts was key to achieving the objective. The finished prototype design uses high-quality materials such as stainless steel, polished aluminum, borosilicate glass, and black walnut wood while balancing a functional design with an open view, showcasing the brew process. Meticulous detail was put into the placement of every component and screw. Finding ways to neatly package the electronics out-of-view was a design challenge.
“Simple can be harder than complex: You have to work hard to get your thinking clean to make it simple. But it’s worth it in the end because once you get there, you can move mountains.” — Steve Jobs
- Notes: My sense was this needed to be designed in up-front, to look nice, and that my ability and tools to work with the materials I have an affinity for such as: stainless steel, copper, glass, walnut wood, could impact the outcome. I would need tools, and knowledge on how to best use these tools with the desired materials.
It needs to be software, and hardware controllable for precision
- At the core of the machine, I developed the software to function as an IoT device, connecting to a home network with support for multiple brew workflows. There are no physical buttons on the machine. I designed the software to be configurable by external interfaces and support a combination of calibration, cleaning, brewing methods with Alexa® control for each brew cycle. The software running on the Arduino hardware is responsible for precision control of components and timings with high-level API’s for web/mobile/Alexa® application use.
- Notes: Despite not knowing exactly how to code for Arduino®, I had a rough idea of the REST api’s I would need to implement on the device. Specific unknowns at this point included how to make an Arduino® process those rest calls. Since this is an area I felt reasonably comfortable with figuring out, I decided I would work on this after the mechanics were in place.
It needs to be operated via web/mobile and Alexa®
- For this goal, I developed an easy-to-use, responsive single page app using React, that would interface with the hardware API’s. This app can work via mobile, or web. However, Alexa® quickly became my favorite means to interface with the machine. It became an incredibly simple way to test the device without opening a browser window.
- Overall this project required 3.5 pieces of software, a JS front-end app, C++ (Arduino®) device-specific software, a broker service (js/node) for the devices to talk to each other, and an Alexa® skill.
- Notes: I would build a simple React Single Page App, and then learn how to build an Alexa® skill that takes voice input/parameters and passes them through to the API’s. I didn’t have experience with Alexa®, but my sense was I’ll figure it out once all the basic API’s were in place.
It needs to be food grade, and easy to clean
- This required ordering custom fit parts made of food-grade materials with temperature considerations to the brew process. Direct contact components included o-rings, valves, seals, filters, metal, glass. Indirect contact materials potentially included wood, which I sealed and coated with food grade oils and wax.
- Notes: I would need to buy food-grade materials and parts that could handle the appropriate temperatures.
It needed to support multiple filters, and work for different grains/grounds
- While the objective here was achieved, I iterated on an improved design that makes changing filters even simpler. I’m personally satisfied with one specific filter, but it can support metal, and cloth filters of various grades.
- Notes: Different mesh filters, materials, all produce variations in taste. I needed to think about the design for this as well as the design tolerances.
It needed to be easily serviceable
- A series of iterations improved the service-ability. Initial designs were simplified and the end-result became something that was much easier to disassemble, and replace individual components. Several 3D printed components were adjusted to better fit the overall machine and simplify the assembly.
- Notes: Having to take apart the machine to fix a single problem would not be ideal.
Each individual goal had a series of hard challenges to overcome that required creative problem solving and refinement. Throughout the process I continuously improved on the materials, iterated on the component design, simplified parts, and improved the electronics until I felt the prototype matched the vision. The push and pull of building a functional, and aesthetically pleasing machine was top of mind throughout the process.
I’m excited to share part III (coming soon!), which will include a look at the prototype, as well as my key takeaways and learnings in drafting/design, electromechanical work, service-ability, machining, and IoT software.
Stay tuned to learn more about the working prototype and the quest to engineer a great cup of coffee. Feel free to reach out via email @ smithxlabs [at] gmail, and follow me on twitter @ smithxlabs
© 2020 Bryan Smith All Rights Reserved.
Disclaimers:
The author assumes no responsibility or liability for any errors or omissions in the content of this site. The information contained in this site is provided on an “as is” basis with no guarantees of completeness, accuracy, usefulness or timeliness.
Apple® is a registered trademark of Apple Inc
Alexa® is a registered trademark of Amazon Inc
If you haven’t read Part I, please check it out here:
